Cyclooxygenase-2 Expression and Its Correlation with Primary Tumor Size and Lymph Node Involvement in Nasopharyngeal Carcinoma.

AIM: This study aimed to observe the cyclooxygenase-2 expression and its correlation with tumour size and lymph node involvement in nasopharyngeal carcinoma.
METHODS: This study was cross-sectional, that enrolled 126 samples diagnosed with nasopharyngeal carcinoma in Haji Adam Malik General Hospital, Medan, Indonesia which fulfilled the inclusion criteria.
RESULTS: Based on this study, we found that the age peak incidence of nasopharyngeal carcinoma patients about a 41-60-year-old group (57.1%), dominated by men (71.4%). Through histopathological examination, non-keratinizing squamous cell carcinoma is the most predominant type (79.4%). We also found T3 is the most prevalent primary tumour size (32.5%) with prominent lymph node involvement N3 (45.2%), and stage IV (54.8%). Cyclooxygenase-2 overexpression is prevalent among nonkeratinizing squamous cell carcinoma (81.1%), T3 primary tumour size (41.1%), N3 node involvement (60.0%), and IV clinical stage (71.6%). In addition, we found a significant relationship between cyclooxygenase-2 expressions towards tumor size (p < 0.001) and lymph node involvement (p < 0.001) in nasopharyngeal carcinoma.
CONCLUSION: It is proved that the overexpression of cyclooxygenase-2 will increase the susceptibility of nasopharyngeal carcinoma patients having advanced primary tumour size and lymph node involvement.

Introduction

Nasopharyngeal carcinoma (NPC) is a malignancy originating from nasopharyngeal epithelial cells [1]. NPC incidence is dominated by Asia population, especially South China and South- East Asia [2]. In Indonesia, there are approximately 6.2 of 100,000 people diagnosed with NPC. Based on the data in 2008, NPC becomes the fifth leading cause of cancer mortality with annual incidence 13,000 [3].

Cyclopentane-fatty acid derivatives, largely known as prostaglandin, are produced in the human cell. It is well-known as a chemical mediator of inflammation. Meanwhile, COX or prostaglandin endoperoxidase synthase is an enzyme involved in the formation of prostaglandin [4], [5]. COX has two isoforms: COX-1 and COX-2. For maintaining homeostasis, COX-1 is expressed by most cells. In contrast, COX-2 expression related to some pathological conditions, particularly in malignancy. COX-2 can induce angiogenesis and inhibit apoptosis. Also, COX-2 has a certain role in resistance to cancer immunotherapy [6], [7], [8], [9]. COX-2 converts arachidonic acid into five major prostanoids: PGE2, PGD2, PGI2 (prostacyclin), PGF2α, and thromboxane A2 (TXA2) [10], [11].

A number of studies have shown that COX-2 will induce angiogenesis by several mechanisms, such as (1) VEGF, it has been explained in the previous section; (2) formation of eicosanoid product (TXA2, PGI2, PGE2), it will directly stimulate endothelial formation, growth, and migration; (3) endothelial cell will be less susceptible to apoptotic by elevation anti-apoptotic Bcl-2 protein expression and activation of the PI3K-Akt pathway; (4) matrix metalloproteinase (MMP) will increase its expression, related to vascular invasion; (5) function of angiogenic protein will increase αvβ3 integrin; (6) IL-12 expression as angiogenesis inhibitor will decrease [2], [12], [13], [14], [15], [16], [17]. In another side the effect of COX-2 on angiogenesis, the effect of COX-2 on lymph nodes involvement still poorly understood but there is an opinion that said macrophages had been suggested as a major source of lymphangiogenic growth factor appeared chronic inflammatory lesions [18].

Our study will show that PGE2 expression, the only one prostanoid, increase as a response to the COX-2 overexpression. COX-2/PGE2 pathway proved its importance in stimulating myeloid-derived suppressor cells production. Myeloid cells can support tumour growth by suppressing immune function and induce angiogenesis [19], [20], [21], [22]. COX or prostaglandin endoperoxidase synthase is an enzyme involved in prostaglandin formation, and its existence is related to inflammation and tumour growth.

We looked at the COX-2 overexpression and compared them COX has two isoforms: COX-1 and COX-2. COX-2 is associated with prostaglandin synthesis in inflammatory tissues and neoplastic processes. In the previous study, COX-2 overexpression is well-documented involved in oncogenesis for certain malignancy, especially NPC [6], [23], [24], [25].

Material and Methods

In this study researchers aimed to observe the cyclooxygenase-2 expression and its correlation with tumour size and lymph node involvement in nasopharyngeal carcinoma. This cross-sectional study enrolled 126 samples diagnosed with nasopharyngeal carcinoma in Haji Adam Malik General Hospital, Medan, Indonesia which fulfilled the inclusion criteria. This study was conducted to analyse the correlation between cyclooxygenase-2 overexpression with tumour size and lymph node involvement in nasopharyngeal carcinoma. It was also expected that it could be used as one of the factors that affect the prognosis in patients with nasopharyngeal carcinoma.

This cross-sectional study conducted in Haji Adam Malik General Hospital and pathology department, Medical Faculty of Universitas Sumatera Utara (USU), Medan, Indonesia. This study enrolled 126 patients diagnosed with NPC by doing anamnesis, physical examination, imaging studies, and histopathological examination.

Then, the samples had to fulfil the inclusion criteria, such as NPC patient diagnosed by histopathological examination and also had never undergone radiotherapy, chemotherapy, or in combination. If the paraffin blocks were not in good condition and the patients were positive with other malignancy, the subject would be excluded from the study. We used non-probability consecutive sampling to avoid tendency and bias and provided the demographic data including gender, age, and some variables related to the samples, condition, including histopathologic type (based on World Health Organization/WHO).

While primary tumour size (T), lymph node involvement (N), and clinical staging were listed based on the American Joint Committee on Cancer (AJCC) 2010 classification. Tissue sections from paraffin block embedded NPC biopsies were stained with Genetex Human COX-2 Antibody. Slides were incubated with the universal peroxidase-labelled polymer (PolyVuePlus) HRP/DAB Detection System and counterstained with hematoxylin. Then, the result classified into four categories by using broad and intensity score, such as 0 = negative, 1 = < 10% of the cell stained or weak-stained, 2 = 10-50% of cells stained or moderate-stained, 3 = > 50% of cells stained or strong-stained. COX-2 immunostaining expression on tissue sections of paraffin block NPC biopsies can be seen in Figure 1.

Figure 1

Strong cytoplasmic expression of COX-2 in non-keratinizing squamous cell nasopharyngeal carcinoma (x 400)

Eventually, the final score was obtained by multiplying broad scores with intensity scores. The score was called as immunoreactive scores, 4 or more defined as positive or overexpression. Histopathological examination and the process related to immunohistochemical staining were done by three professional pathologists.

The statistical analysis was done by using SPSS 16.0 software (SPSS, Chicago, IL). Demographic data were listed in a univariate variable. Then, bivariate analysis was performed using chi-square (χ2) test to determine the relationship between COX-2 overexpression with primary tumour size and lymph node involvement in NPC. This study also had been approved by the Health Research Ethical Committee of Medical Faculty of Universitas Sumatera Utara, Medan, Indonesia.

Results

This study was conducted on 126 NPC patients who fulfilled the inclusion criteria. The highest prevalence of NPC was found in the age group of IV to VI decades (57.1%), dominated by males (71.4%) with non-keratinizing squamous cell carcinoma (79.4%), primary tumor size T3 (32.5%), lymph node involvement N3 (45.2%), and clinical stadium IV (54.8%). Based on immunohistochemical examination overexpression of COX-2 were also found in nonkeratinizing squamous cell carcinoma, primary tumour size T3, lymph node involvement N3, and clinical stadium IV. Based on the results of this study, we found a significant relationship between COX-2 expression with tumour size and enlarged lymph nodes with a value of p < 0.001.

Discussion

In our study, a positive correlation between COX-2 overexpression with primary tumour size and lymph node involvement was proved in NPC. It explains COX-2 overexpression is more common in advanced primary tumour size and lymph node involvement. Bin Yang et al., also discovered the same result related to COX-2 overexpression and malignant process, including metastasis [26]. A study conducted by Gui Yang et al., also found that COX-2 and advanced clinical stage of NPC are positively correlated (OR 5.39; 95% CI: 3.79-7.66) [27]. Also, One study conducted by Li et al. found by using nasopharyngeal carcinoma cell lines treated with nonsteroid anti-inflammatory drugs (NSAIDs), particularly celecoxib, invasion, and migration will decrease through suppression of MMP-2 and -9 activity 28 Fendri et al., also discovered the similar result related to COX-2 overexpression that There was a significant association between COX-2 expression with lymph node involvement (N+) in NPC patients with p < 0.0001 [29].

Besides VEGF, Epidermal growth factors (EGF) also play an important role in tumour proliferation and invasion. Ross found that COX-2 overexpression occurred in 79% of NPC patients, and it related to EGFR status but not with latent membrane protein (LMP) -1 or inducible NOS. Meanwhile, Tan K-B prevailed that COX-2 might be involved in the multistep process of NPC carcinogenesis since the COX-2 expression is more common in a dysplastic nasopharyngeal cell [30], [31]. It is also stated by Kwong et al. which enrolled 53 NPC patients, all patients who had intense staining for COX-2 were dysplastic [32]. Many clinical studies have shown that COX-2 induces angiogenesis, but there is no evidence to prove the relationship between COX-2 and lymph node involvement although many previous studies suggested that expression of COX-2 correlated with lymph node metastasis [18].

Also, a prognostic study conducted by Pan et al. uncovered that survival rate, including overall survival, disease-free survival, locoregional control, and distant metastasis-free survival are also related to COX-2 overexpression. Chen et al. conducted a study in T4 NPC patients treated with radiation therapy, 5-year survival rates for patients who have COX-2 overexpression was 27% compared with 60% in low COX-2 expression group. Furthermore, Xinhua et al., also stated that COX-2 might be used to predict prognosis, particularly local recurrence and distant metastasis [33], [34], [35]. Prognostic study related to COX-2 overexpression was evident in many cancer one of them is glottis cancer, COX-2 overexpression related to more aggressive tumour and low survival rate [36], [37], [38]. Otherwise, Long et al., and Y. J. Kim et al. found a contradictive result in our study. By using smaller sample sizes, both studies concluded that COX-2 expression and tumour size did not correlate [39], [40].

Our study provided evidence that the correlation between COX-2 overexpression with primary tumour size and lymph node involvement are significant (p < 0.001). The COX-2 expression will increase as the tumor size increases. COX-2 overexpression is one of the tumor lymphangigenesis factors. Also, COX-2 contributes in the process of carcinogenesis, and it allows COX-2 to be used as a therapeutic target in the future for nasopharyngeal carcinoma.

Table 1

COX-2 overexpression frequency based on histopathology type, lymph node involvement (N) frequency, and clinical staging

Type		COX-2 Expression
Over-expression		%	Negative	%	P**
Histopathology					0.253
Keratinizing SCC*	7	(7.4)	1	(12.5)
Non-keratinizing SCC*	77	(81.1)	23	(23)
Undifferentiated carcinoma	11	(11.6)	7	(38.9)
Primary Tumor Size (T)					0.000
T1	5	(5.3)	20	(64.5)
T2	24	(25.3)	8	(25.8)
T3	39	(41.1)	2	(6.5)
T4	27	(28.4)	1	(3.2)
Lymph node involvement (N)					0.000
N0	9	(9.5)	18	(58.1)
N1	14	(14.7)	5	(16.1)
N2	15	(15.8)	8	(25.8)
N3	57	(60.0)	0	(0.0)
Clinical staging					0.000
I	2	(2.1)	15	(48.4)
II	7	(7.4)	9	(29.0)
III	18	(18.9)	6	(19.4)
IV	68	(71.6)	1	(3.2)

SCC: Squamous Cell Carcinoma

p: P-value